Methodology ofMagnetic Resonance Spectroscopy:

MRS

Graeme Mason, Ph.D.

Depts. of Psychiatry and Diagnostic RadiologyYale University, New Haven, CT

Welcome to New Haven

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

Methods: What is MRS?

Nucleus

Electron(s)

Spin of nucleus is either ‘up’ or ‘down’.

Nuclear Magnetic Resonance Spectroscopy

Methods: What is MRS?Subject lies inside a large magnet to orient the spins.

The floor

The Magnet

The

Wal

lTh

e Do

g

Rotating Magnetic Dipole

Nucleus N

S

Bo

In a large magnetic field (scanner magnet), the magnetic dipole precesses about the axis of the large field with a frequency ν = γBo/2π.

Only nuclei with spin can be detected with MRS.

Some NMR-Visible Nuclei1H – water, lipids, amino acids, many other metabolitesγ = 2.67 x 104 radians/sec/Gauss

13C - amino acids, neurotransmitters, glucose, lipids, acetateγ = 0.67 x 104 radians/sec/Gauss

15N - metabolism of ammonia, amino acids

19F - pharmacokinetics of fluoxetine and fluvoxamine

23Na - effects of hypoxia, challenges to Na pumps

2H - metabolism of fats

7Li - pharmacokinetics of lithium

Transition from Quantum to Classical Mechanics1.5 mM GABA, 13.5 cc voxel contains 2.4 x 1022 hydrogen nuclei for GABA detection

The large number of particles allows a meaningful analysis of the behavior of the whole group.

Boltzmann distribution:N+/N- = exp(-γhBo/2πkT)

approximated asN+/N- = 1 - γhBo/2πkT

N+ minus N- is the net magnetization:

Spin Orientation in a Head(Outside the Scanner)

Spins evenly mixed (N+ = N-)

=

Bo Spins

Inside Scanner(Strong Magnetic Field)

0.00001 - 0.00004% align with the magnetic field,creating net magnetization in the head.

NetMagnetization

xy

z

What is MRS?Subject lies inside a large magnet to orient the spins.

The Magnet

The RF Coil

=Bo

RadioFrequency

xy

z

Radio Frequency Energy Applied

The net magnetization is altered by the application of energy at the proper frequency.

=Bo

xy

z

Longitudinal (T1) Relaxation

The magnetization returns to its original orientation at a rate governed by the exponential constant T1.

RadioFrequency

Transverse (T2) Relaxation

Phase coherence is lost as the spins dephase due to microscopic magnetic field inhomogeneity.

xy

z

xy

z

xy

z

z

xy

xy

NetNet

1 2 3

SIGNALSIGNAL SIGNAL

MRS Frequencies Distinguish Chemicals:Sources of Differences in NMR Frequencies

1. Nucleus and Magnetic Field Strength (MHz)Nucleus:Example: 13C frequency ≈ 25% 1H frequency

Magnetic Field Strength:For 1H, 42.6MHz x field strength (Tesla)Examples:

a. 1.5T = 63.9 MHzb. 2.1T = 89.4 MHz (~public radio - Bridgeport)c. 4 T = 170 MHz

Sources of Differences in NMR Frequencies2. Molecular identity (Hz to thousands of Hz)

GlutamineCOO-

NH3+

CCC

H2

NH2HH2

COO-

COO-

CCC

H2

NH2HH2

Glutamate- Different molecules respond with slightly differentfrequencies.

- At 2.1T, glutamate responds at 89,633,931 Hzglutamine responds at 89,633,940 Hz

3. J-coupling: (Hz to hundreds of Hz) (used for GABA editing)

1.52.02.53.03.54.0ppm

Cr/PCr NAAGln C4

Cho

Glu C4Glx C3

NAA

MRSMRS of the human brain: chemicals separated by frequency

1.52.02.53.03.54.0ppm

Cr/PCr NAAGln C4

Cho

Glu C4Glx C3

NAA

MRS: Significance

Glutamate, Glutamine, Glx: neurotransmission, energetics

Gln C4

Glu C4Glx C3

N-AcetylAspartate: neuronal viability, energetic link?

NAA

NAACr/PCr: Total of Creatineand Phosphocreatine:Largely inert

Cr/PCr

Choline:Membrane Turnover?Cell Damage? Cell Repair?

Cho

1.52.02.53.03.54.0ppm

Cr/PCrNAA

Gln C4

Cho

Glu C4Glx C3

NAA

MRSShort-echo measurement of glutamate, N-acetylaspartate, and other metabolite: What else is present?

1.52.02.53.03.54.0ppm

Macromolecules

Total Signal

MRSShort-echo measurement of glutamate, N-acetylaspartate, and other metabolite: Macromolecules in baseline

Cr/PCrNAA

Gln C4

Cho

Glu C4Glx C3

NAA

1.52.02.53.03.54.0ppm

Cr/PCrNAA

Gln C4

Cho

Glu C4Glx C3

NAA

Difference:Just Metabolites

Macromolecules

Total Signal

MRSShort-echo measurement of glutamate, N-acetylaspartate, and other metabolite: Macromolecules out. What else is present?

1.52.02.53.03.54.0ppm

Cr/PCrNAA

Gln C4

Cho

Glu C4Glx C3

NAA

GABA(buried)

Difference:Just Metabolites

Macromolecules

Total Signal

MRSShort-echo measurement of glutamate, N-acetylaspartate, and other metabolite: Measurement and Removal of Macromolecules

MRS1H MRS measurement of GABA, Bruker 2.1T MR scanner, 13-22 cc occipital voxel: J-editing difference method

-1.01.0ppm5.07.0

Cho

GABA

CrWater

Difference x 4

NAA

Subspectrum

3.0

GABA Editing

GABA

CreatineCholine

Glutamate+Glutamine

•8 cc at 4T.

0123456 ppm

GABATime 1

Time 2

GABA in One Alcohol-Dependent Subject

Spectra of cortical GABA obtained in one subject at one week and one month of sobriety. GABA was reduced at one month.

0.0

0.4

0.8

1.2

1.6

2.0

[GA

BA

] (m

mol

/kg)

Time 1 Time 2 Time 1 Time 2

Non-smokers Smokers

-0.80

-0.60-0.40

-0.20

0.00

0.200.400.60

Cha

nge

in [G

AB

A] (

mm

ol/k

g)

•Changes in GABA levels with sobriety depended on smoking (p = 0.03).

•GABA in non-smokers fell by 0.39 ± 0.18 mmol/kg (p = 0.004), but GABA in smokers did not (p = 0.75).

•GABA in non-smokers at time 1 differed from the non-smokers’ GABA at time 2 (p = 0.03).

Effects of 1 Month of Sobriety on GABA

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

1H MR Spectroscopic Imaging TE = 25 ms Metabolite Distribution and Quantification

Soher B et al. MRI 2000

T1w NAA

ChoCr myo-Ins

2.0PPM 3.0

B

A

2.0PPM 3.0

Data & Base

Data - Base&

Metabolites

Data & Base

Data - Base&

Metabolites

mI Cho Cr NAAMRI SI

spatial resolution 1 ml

A

B

SI slicepositions

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

Image Segmentation

Creating images to quantify GM, WM, and CSF

%WM%GM

%CSF

Image Segmentation

Volumes expressed as % of intracranial volume (ICV)

Meyerhoff, ICANA 2004

Correlations of Gray, White, and Total Tissue Volumesin Alcoholism

•Across timepoints, there was a significant correlation between white matter content in the voxel and the total solid tissue (p < 0.0005), but not between grey matter and solid tissue (p = 0.36).

•Significance: long-term improvements in occipital tissue composition are primarily from white matter.

10

20

30

40

50

60(%

of M

RS

voxe

l)

0

10

20

30

40

50

75 80 85 90 95 100Tissue (% of MRS voxel)

(% o

f MR

S vo

xel)

Grey Matter

White Matter

Short-term sober subjects

Long-term sober subjects

Alcohol-Related Brain Atrophy

How Can Segmented Data Be Useful?

When pure tissue is not accessible, regression analysis can be used to estimate concentrations in pure tissue.

Chu et al., Magn Reson Med, 2000

How Can Segmented Data Be Useful?

A disease state can be examined with MRSI and image segmentation. Statistically abnormal pixels can be highlighted.

Chu et al., Magn Reson Med, 2000

How Can Segmented Data Be Useful?

When pure tissue is not accessible, regression analysis can be used to estimate T2 effects in pure tissue.

Sammi et al., Magn Reson Med, 2000

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

13C MRS:Rates of Metabolism and Neurotransmission,

Clear Resolution of Glutamate and Glutamine

• 13C non-radioactive isotope of carbon.

• 1% natural abundance means low background.

• No isotope effect on metabolism.

• Natural labeled substrates in blood yield MRS-detectable products in glutamate, glutamine, and other compounds in the brain.

13C Isotopic Labeling

[1-13C]Glucose

[4-13C]Glutamate [4-13C]Glutamine

Time courses of labeling yield rates of metabolic pathways.

140 cc voxel, 45 min accumulation

5152535455565758595105ppm

Glucose C1 αGlu C2

Gln C2Cr C2Cho C2

NAA C2Asp C2

GABA C4+ NAA C3 Asp C3

Cr-CH3

Glu C4Glu C4,3+GABA C2 Glu C4,3

Gln C4

lipid

Glu C3,4Glu C3

Glu C3,4+Gln C3 NAA C6

Lac C3

[1-13C]Glucose Administration

Natural Abundance

Glucose C1 β

A 13C NMR Spectrum of a Human Brain in Vivo

Isotopic Flow

•Faster flow more rapid appearance of dye at each pool

•Precursor pools “trap” dye temporarily

Glutamate

Glucose

Lactate

Glutamine

Hana, Maui

Yellow Dye

13C Labeling of Neuronal Glutamate & Astroglial Gln by Glutamate/Glutamine Cycle

•

Vcycle

Gln4*ECF

Glu4* ECFGlu4* ECF

Gln4*Gln4*

Glu4*Glu4*

Gln4

TCAA

VTCA

TCAN

VVTCATCA

Vefflux

OAA

Glu4*Glu4*

Vcycle

αKG4

Astrocyte Neuron

Blood Brain

PDH

[1-13C]Glucose Pyr3Pyr3

KG4*

Vcycle

KG3*

Glu3* ECFGlu3* ECF

Glu3*Glu3*

Gln3*Gln3* Gln3*ECFGln3

αKG3

Vx

13C Labeling of Neuronal Glutamate & Astroglial Gln by Glutamate/Glutamine Cycle: VX, Vtca, Vcycle

0

10

20

30

40

50

60

0 50 100 150

Gln C4 PE(%)Gln C4 DataGln C3 PE (%)Gln C3 Data

0

10

20

30

40

50

60

0 50 100 150

GluTot4 C# PE(%)GluTot4 C# DataGluTot3 C#Data

CMRgl(ox) = 0.13 + 0.96Vcycle, R2 = 0.74

Determining a Relationship Between Glutamate Release and Glucose Oxidation

~1:1 relationship between changes in glucose oxidation and glutamate-glutamine neurotransmitter cycling

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Vcycle (µmol/min/g)

CM

Rgl

(ox)

(µm

ol/m

in/g

) Predicted

Morphine

Morphine + Nicotine

Morphine + Pentobarbital

Isoelectric Pentobarbital

a-Chloralose

13C MR Time Courses of Glu and Gln in Individual Patients and Controls

0.0

5.0

10.0

15.0

20.0

25.0

30.0G

luC

4 En

richm

ent (

%)

0.0

5.0

10.0

15.0

20.0

25.0

0 50 100 150time (min)

Gln

C4

Enric

hmen

t (%

)

Depressed

Control

Depressed

Control

Glutamate

Glutamine

Glutamate-glutamine cycling is reduced in depression.

No difference in oxidative metabolism.

13C MRS Application to Human Disease:Measurement of NAA Synthesis in Canavan Disease

Moreno et al., J Neurochem, 2001

13C MRS Application to Human DiseaseGrading of Hepatic Encephalopathy

Blüml et al., Magn Reson Med, 2001

Overview

What is MRS?

What is Spectroscopic Imaging?

What is Image Segmentation?

What is 13C MRS?

Acknowledgements

Alcohol StudiesJohn Krystal (CTNA director)Ismene PetrakisLouis TrevisanElizabeth RuffVladimir CoricRalitza Gueorguieva

MRSDouglas Rothman (MRRC director)Kevin Behar (GABA expertise, director mouse imaging program)Robin de GraafMichael AppelNicola Sibson

Acknowledgements

R01 AA 11321-01A2I-P50 AA-12870-01K02-AA13430-01

NIAAA Center for theTranslational Neuroscience

Of Alcoholism

NIAAADept. Veterans Affairs